The prognostic role of cardiac troponin in hospitalized COVID-19 patients

Myocardial injury defined by elevation of cardiac troponins (cTn) in the course of coronavirus disease 2019 (COVID-19) pandemic has been reported, though not fully characterized yet. Using the Turkish nationwide centralized COVID-19 database, we sought to determine whether cTn measured within 24 h of admission may help identify 30-day all-cause mortality in hospitalized patients. This retrospective cohort study was conducted at all hospitals in Turkey between March 11, 2020, and June 22, 2020. All hospitalized COVID-19 patients (≥18 years) who had cTn measurements within 24 h of admission were included. The primary outcome was 30-day all-cause mortality. A total of 14,855 COVID-19 patients (median age 49 years and 54% male) from 81 provinces of Turkey were included. Of these, 2020 patients (13.6%) were transferred to intensive care unit, 1165 patients (7.8%) needed mechanical ventilation, and 882 patients (5.9%) died during hospitalization. The prevalence of cTn positivity was 6.9% (n = 1027) in the hospitalized patients. cTn positivity was 5% in those patients alive at 30-day, and 44% in those who died. In multivariable Cox proportional hazard regression model, age, lactate dehydrogenase, and cTn were the strongest predictors of 30-day mortality, irrespective of cTn definition as a continuous, ordinal variable, or dichotomic variables. A single measurement of cTn at admission in patients with COVID-19 is associated with 30-day all-cause mortality and may have an important prognostic role for optimizing risk stratification. [Display omitted] •A total of 14,855 hospitalized COVID-19 patients who had cardiac troponin (cTn) measurement within 24 h of admission were included in the study, and cTn positivity rate was 6.9%.•Even small amounts of cTn elevation were significantly associated with 30-day mortality, while greater amounts of cTn elevations were significantly associated with higher risk of 30-day mortality after adjustment for clinically relevant factors.